We aim to enhance the ability to design protein crystallographic constructs and, thereby, substantially speed protein structure determination with the use of information provided by innovative peptide amide hydrogen exchange techniques. In this resubmission, we present extensive preliminary studies, performed since review, that directly address all reservations regarding our prior submission. Determination of high-resolution structure at an increasingly high throughput (HT) pace is required for a fundamental understanding of how modifications of cancer- implicated proteins can promote oncogenesis and metastasis. Unfortunately, HT crystallographic efforts have a single, dominating roadblock: they produce suitable crystals for a small minority of target proteins. Floppy, unstructured regions of failed proteins play a major role in this problem. The exchange rates of the many peptide amide hydrogens within a protein are determined by the protein's stability at the individual amino acid scale. We have developed an enhanced form of amide hydrogen/deuterium exchange-mass spectrometry (DXMS) that can rapidly and precisely measure such rates. We propose that DXMS data can be used to identify and localize such unstructured regions within a protein and thereby guide the design of modified protein in which such regions are selectively removed. Furthermore, many proteins require tertiary-quaternary contacts, provided by binding partners, to induce structure in such regions. For these proteins, DXMS can be used to rapidly select binding partners that provide the needed stabilizing contacts, allowing focused protein-binding partner co-crystallization efforts. Importantly, repeat DXMS study of the modified protein(s) can rapidly determine how well they have retained the structured elements of the original protein. In our R21 year, we will demonstrate that DXMS can guide the re-design of protein constructs sufficiently to produce a 50% increase in overall crystallization success rates for target proteins, and do this at a high throughput pace. This result will establish the ability of DXMS to speed throughput of present HT crystallographic efforts, and likely similarly enhance construct definition for conventional, specific-protein focused crystallography, with obvious benefits for the structural study of cancer related proteins. In our R33 years we will establish a crystallography-dedicated DXMS facility and further refine our ability to guide construct design by analysis of the protein targets studied by our collaborators at the Joint Center for Structural Genomics, with an emphasis on those with cancer-relevance. This construct-refinement resource will then be broadly extended as a community service to NCI-funded investigators for application to both conventional and HT crystallographic efforts.